Upper case the console messages and stamp the banner (#98) - #99
Merged
Conversation
The operator console output was mixed case, named a version hardcoded in
ftpd.h, and reused message IDs for unrelated events. All three are now
aligned with UFSD.
Build stamp. Bump the mbt submodule to mbt#64 so .mbt/buildstamp.h is
generated, and use it for the banner:
FTPD000I FTPD 1.0.0-DEV (CF14987) STARTING
FTPD005I LIBC370 1.0.2-DEV (5C0DEEB)
FTPD006W BUILT FROM A MODIFIED WORKING TREE
...
FTPD001I FTPD 1.0.0-DEV READY
A deploy/relink mismatch -- sysroot says X, the STC runs Y -- can no
longer hide, and a build carrying uncommitted tracked changes says so
instead of passing itself off as the commit it was branched from.
buildstamp.h is included per translation unit (ftpd.c, ftpd#con.c,
ftpd#cmd.c) and deliberately not from ftpd.h: -MMD would otherwise make
a new commit rebuild every object. FTPD_VERSION and FTPD_VERSION_STR
are gone; FTPD_PRODUCT holds the name the STAT reply still needs.
Case. Every ftpd_log_wto() message is upper case; values -- data set
names, VOLSERs, the banner, every %s/%d substitution -- keep their
original case. ftpd_upcase() (ftpd#log.c) handles the three strings
that arrive lower case at runtime: MBT_VERSION, MBT_COMMIT and
libc370_version(). It uses the libc370 toupper(), so the mapping is
EBCDIC-correct.
ftpd_session_reply() is untouched: those are FTP protocol replies read
by clients, not console output. So is ftpd_log(), which goes to SYSOUT.
Message IDs. Six were ambiguous or in the wrong slot:
FTPD097I shutting down -> FTPD098I
FTPD098I shutdown complete -> FTPD099I
FTPD099E COM area not available -> FTPD090E (was sharing 099E)
FTPD099E initialization failed -> FTPD091E (was sharing 099E)
FTPD090E unknown command / TRACE -> FTPD021E (UFSD's 020I/021E pair)
FTPD005W SSLPROXY (x2) -> FTPD007W (005 is libc370 now)
The two APF success messages are gone -- UFSD reports only the failure,
and one of them was a second FTPD000I competing with the banner. The
warn-and-continue behaviour on APF failure (FTPD003W) is unchanged.
/F FTPD,VERSION now repeats the banner identity instead of a static
string, which is what an operator asking it after a deploy wants.
MCS prefixes every message from an unauthorized problem program with
'+'. clib_apf_setup() ran inside initialize(), i.e. after the banner
had already gone out, so the startup showed one line marked up
differently from all the others:
+FTPD000I FTPD Server 1.0.0-dev starting
FTPD000I FTPD was APF authorized via SVC 244
Hoist the call into main() between __cibset() and the banner, which is
where UFSD has it. The rc is carried past the banner rather than
reported at the call site: a failure is worth a line, but not one ahead
of the message saying which server is starting. FTPD still warns and
continues where UFSD gives up -- that behaviour is unchanged.
initialize() no longer needs argc/argv and drops both parameters.
Contributor
Author
|
Follow-up in db28e70: the banner also had a MCS prefixes every message from an unauthorized problem program with The SOLL has no
Also ran the test suites, since the mbt bump carries two test-infrastructure |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #98.
Before / after
Plus, only when the module was built from a tree with uncommitted tracked
changes (mirrors
UFSD006W):What changed
Build stamp.
mbtbumpedbf0e081->8b8a2f0(mbt#64, fast-forward) so.mbt/buildstamp.his generated. The banner now carriesMBT_VERSION,MBT_COMMITandlibc370_version(), so a deploy/relink mismatch cannot hide.buildstamp.his included per translation unit (ftpd.c,ftpd#con.c,ftpd#cmd.c) and deliberately not fromftpd.h---MMDwould otherwise makea new commit rebuild every object.
FTPD_VERSION/FTPD_VERSION_STRare gone;FTPD_PRODUCTholds the name the STAT reply still needs.Case. Every
ftpd_log_wto()message is upper case; values (data set names,VOLSERs, the banner text, every
%s/%d) keep their original case.ftpd_upcase()inftpd#log.chandles the three strings that arrive lower caseat runtime -- it uses the libc370
toupper(), so the mapping is EBCDIC-correct.Message IDs.
FTPD097Ishutting downFTPD098IFTPD098Ishutdown completeFTPD099IFTPD099ECOM area not availableFTPD090EUFSD090E; was sharing 099EFTPD099Einitialization failedFTPD091EFTPD090Eunknown command / TRACE syntaxFTPD021EFTPD005WSSLPROXY (x2)FTPD007WThe two APF success messages are gone -- UFSD reports only the failure, and one
of them was a second
FTPD000Icompeting with the banner. The warn-and-continuebehaviour on APF failure (
FTPD003W) is unchanged./F FTPD,VERSIONnow repeats the banner identity (version, commit, libc370)instead of a static string.
Out of scope
ftpd_session_reply()was not touched -- those are FTP protocol replies read byclients, not console output -- nor
ftpd_log(), which goes to SYSOUT.Verification
makeunder-Wall -Werror, no warnings.FTPDnnnXliteral:all upper case, no duplicate IDs,
LIBC370 1.0.2-dev (5c0deeb)linked in.#if MBT_COMMIT_DIRTYchecked both ways:FTPD006Wis in the module whilethe tree carries the uncommitted submodule bump, and compiled out after the
commit (
MBT_COMMIT "ca2b42d",MBT_COMMIT_DIRTY 0).doc/FTPD_RAKF_SETUP.mdupdated for theFTPD004I/FTPD004Wwording.The console output itself is unverified until the STC is restarted on the
target -- the checks above are on the built module, not a running server.